Practical AI workflows and tutorials for busy professionals. Plus weekly news analysis, without the noise.
Live Last read · last published · next check
This week in AI: Anthropic's Claude cracks open a 167-year-old maths problem the same week it watermarks every future output, a Google DeepMind model lets Deaf users sign to their Pixel 11, and xAI's Grok Bot works your inbox while you sleep.
Your brain mistakes attention for connection.
This week in AI: Google's leadership shake-up as Hassabis steps back and Jeff Dean departs, Hark Labs' Handoff agent shops with real credit cards, and ByteDance's 30-second video model with a 10-trillion-parameter run behind it.
The five-minute setup that puts your brand on everything Claude Design makes.
This week in AI: DeepSeek undercuts OpenAI with an open-weights model at $0.14 per million tokens, Google DeepMind gives humanoid robots whole-body control, and AI music generator Suno loses its copyright battle in Munich.
Claude can now learn a task by watching you do it once.
This week in AI: Anthropic releases Opus 5 and Claude learns by watching you work, OpenAI puts voice in charge of your computer, and Alibaba's Qwen 3.8 claims frontier scale with open weights.
The deep irony in consulting AI to decide whether a human wrote something, because AI is evil.
This week in AI: Your AI double arrives in Gemini, Anthropic's $10B compute deal with Meta, and Thinking Machines releases a free frontier-scale model.
Where the randomness comes from and how to make it work for you.
